Preoperative or neoadjuvant chemotherapy is now the standard of care for treatment of patients with locally advanced breast cancer and is being evaluated in patients with earlier stage disease. Neoadjuvant chemotherapy can achieve tumor downstaging and allow breast conservation in patients for whom initially, mastectomy was the only option. Response to chemotherapy is assessed clinically and clinical assessment of response of the primary tumor has been shown to be associated with improved disease-free survival. Nonetheless, clinical response does not accurately reflect pathologic response. MRI can accurately assess the extent of cancer in the breast and may be more effective than clinical exam at measuring changes in tumor size and distribution in response to neoadjuvant chemotherapy. Additionally, early tumor changes measured by MRI could be meaningful predictors of survival. More importantly, if MRI can identify those patients who are unlikely to respond to treatment, a change in management can be introduced at an earlier time. This study proposes to investigate these possibilities by using MRI to non-invasively measure tumor changes in patients with Stage III/IV breast cancer during neoadjuvant treatment. The goal is to design and validate MRI methods for measuring treatment response and predicting patient outcome. A high spatial resolution three time-point method was previously developed and evaluated for characterization of breast disease. This technique will be used to measure morphologic and enhancement properties of tumors and monitor their change over treatment. The effectiveness of MRI methods for measuring tumor response and predicting disease-free survival will be investigated in a group of 75 patients with locally-advanced breast cancer who are undergoing neoadjuvant chemotherapy. Clinical investigations will be supported by studies in model systems. Experimental models of breast cancer will be used to study the effects of tumor properties and treatment parameters on treatment response. These results will be used to guide the design of clinical studies. Experimental model studies will be used to study the anti-angiogenic properties of anti-VEGF and anti-FLKI, both being introduced into clinical trials for breast cancer. A third agent to be studied is an anti-body labeled immunoliposome agent containing doxorubicin and targeted against HER2, which has demonstrated superior therapeutic results to both free doxorubicin and non-targeted doxorubicin-containing liposomes. MRI methods will be developed using both standard gadolinium-DTPA and gadolinium-encapsulating immunoliposomes.